Why ‘average’ returns aren’t good enough Commentary: The futility of using past market data for predictions. It has been a tumultuous couple of weeks in the stock and bond markets. But nonetheless, Wall Street wants me to relax. Feel good. Look on the bright side. Stop worrying. The brokerage firms and investment banks have their Ph.D.s and their CFAs and their PowerPoint presentations. Mahogany desks. Neutral tones. Comfortable sofas. As the stock market spirals higher, financial advisers and portfolio managers from Long Beach to Long Island are telling clients much the same thing. The conventional wisdom goes something like this: “Stocks and bonds follow what is known as a random walk. No one can ever know in advance what the returns will be next month or next year. However, over the longer term we have better information. We have several decades’ worth of past data. Since 1962, for example, U.S. stocks have produced average returns in a typical year of 11% and U.S. Treasury bonds about 7%. So a balanced portfolio of 60% stocks, 40% bonds produced returns in the average year of about 9.5%. We also know the standard deviation of annual returns. Based on this we can estimate returns over any five- to 10-year period with 95% or even 99% confidence. And the more “risk” or volatility you are willing to accept over shorter periods, the higher your returns over longer ones. Those who wish or need to earn more can just ramp up the risk, like turning up the volume on your hifi.” It’s a very compelling sounding argument. Those offering it often mention a few Nobel Prize-winning economists to back it up. They sound very sure of themselves. And I wish I could just relax and believe them. I want to. I really, really do. But every time I hear arguments like this — and I hear them a lot — I keep getting a few niggling doubts. There is, for example, the worrying reliance on “averages.” You can do a lot with averages. Bill Gates and I have an average net worth of about $20 billion. Lucky me. Albania and China have an average population of about 600 million people. Telling me about “average” stock market returns poses as many questions as it answers. Oftentimes, too, money managers cite arithmetic rather than geometric averages. This may sound like a minor technical issue, but it isn’t. If the stock market halves, and then doubles, I am merely back where I started. I have made no money at all, and have had enormous numbers of sleepless nights for my pains. My “geometric” return is 0%. But my two annual returns are -50% and +100%. So according to “arithmetic” returns, I’ve just earned average returns of 25% a year. I notice, worryingly, that a lot of Wall Street marketing materials seem to quote those arithmetic averages. Is it a coincidence that they produce better-looking numbers? Since the early 1960s, for example, the Standard & Poor’s 500 index has produced arithmetic returns of 11.1% a year, according to the NYU Stern School of Business. But the geometric average was 9.7% — almost one and a half points lower per year. Another thing that worries me is that these returns are often quoted without taking into account inflation. That’s a story for another day, but if I make a 5% return, and my costs go up 6%, have I done well or badly? Another thing worrying me is this extraordinary reliance on a few decades’ data. Back in the day, when I first became an analyst at a big strategic consultancy, I looked around me and noticed I was the only humanities graduate in the room. Consultancies and Wall Street banks hire tons and tons of math and science Ph.D.s, all of whom are very good at building complex spreadsheets, but they hire very few historians. Those who have studied history know that it is full of discontinuities. You can’t predict the future by clicking and dragging from the recent past. Silahkan daftar akun baru: ironfxsolidtradingindonesia.blogspot/
Posted on: Fri, 28 Jun 2013 11:47:51 +0000
Trending Topics
Recently Viewed Topics
© 2015